Meta-Llama-3-8B-Instruct

24 records across 2 hardware.

hardwaremodelquantbackendtok/spp tok/sWttft sctx/tgdateclassflagssource
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–310.24–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–318.84–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–362.25–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–426.41–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–169.92–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–406.99–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–181.72–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–454.36–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–197.86–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–531.68–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–204.92–––2026-10-03community–source
MI50Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–571.23–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–69.09–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–92.57–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–98.26–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–117.49–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–128.82–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–128.20–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–135.69–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–135.13–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–191.79–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–188.99–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–204.13–––2026-10-03community–source
gfx1152Meta-Llama-3-8B-Instruct Q2_Kllama.cpp (ROCm)–205.85–––2026-10-03community–source

← all models